High-level production of valine by expression of the feedback inhibition-insensitive acetohydroxyacid synthase in Saccharomyces cerevisiae.

全文

(1)1. High-level production of valine by expression of the feedback inhibition-insensitive. 2. acetohydroxyacid synthase in Saccharomyces cerevisiae. 3 4. Natthaporn Takpho, Daisuke Watanabe, Hiroshi Takagi*. 5. Graduate School of Biological Sciences, Nara Institute of Science and Technology, 8916-5. 6. Takayama, Ikoma, Nara 630-0192, Japan. 7 8. *. 9. and Technology, 8916-5 Takayama, Ikoma, Nara 630-0192, Japan. E-mail address:. 10. Corresponding author: Graduate School of Biological Sciences, Nara Institute of Science. hiro@bs.naist.jp (H. Takagi). 11 12. Keywords: valine; yeast; acetohydroxyacid synthase; Ilv6; feedback inhibition; branched-. 13. chain amino acids.. 14 15. Abstract. 16. Valine, which is one of the branched-chain amino acids (BCAAs) essential for humans, is. 17. widely used in animal feed, dietary supplements and pharmaceuticals. At the commercial level,. 18. valine is usually produced by bacterial fermentation from glucose. However, valine. 19. biosynthesis can also proceed in the yeast Saccharomyces cerevisiae, which is a useful. 20. microorganism in fermentation industry. In S. cerevisiae, valine biosynthesis is regulated by. 21. valine itself via the feedback inhibition of acetohydroxyacid synthase (AHAS), which consists. 22. of two subunits, the catalytic subunit Ilv2 and the regulatory subunit Ilv6. In this study, to. 23. improve the valine productivity of yeast cells, we constructed several variants of Ilv6 by. 24. introducing amino acid substitutions based on a protein sequence comparison with the AHAS. 25. regulatory subunit of E. coli. Among them, we found that the Asn86Ala, Gly89Asp and. 1.

(2) 26. Asn104Ala variants resulted in approximately 4-fold higher intracellular valine contents. 27. compared with those in cells with the wild-type Ilv6. The computational analysis of Ilv6. 28. predicted that Asn86, Gly89 and Asn104 are located in the vicinity of a valine-binding site,. 29. suggesting that amino acid substitutions at these positions induce conformational change of the. 30. valine-binding site. To test the effects of these variants on AHAS activity, both recombinant. 31. Ilv2 and Ilv6 were purified and reconstituted in vitro. The Ilv6 variants were much less sensitive. 32. to feedback inhibition by valine than the wild-type Ilv6. Only a portion of the amino acid. 33. changes identified in the E. coli AHAS regulatory subunit IlvH enhanced the valine synthesis,. 34. suggesting structural and/or functional differences between the S. cerevisiae and E. coli AHAS. 35. regulatory subunits. It should also be noted that these amino acid substitutions did not affect the. 36. intracellular pools of the other BCAAs, leucine and isoleucine. The approach described here. 37. could be a practical method for the development of industrial yeast strains with high-level. 38. production of valine or isobutanol.. 39 40. 1. Introduction. 41. Valine is one of the branched-chain amino acids (BCAAs) essential for humans, along with. 42. leucine and isoleucine, and valine deficiency has been linked to several human diseases. 43. (Hutchison et al., 1983; Burrage et al., 2014). Since valine is a hydrophobic amino acid, it often. 44. forms the helical structures within interior proteins. Valine is included in various commercial. 45. products, such as animal feed, human dietary supplementaries, pharmaceuticals and cosmetics.. 46. In addition, derivatives of valine can be used as a substrate for antibiotics and antivirals as well. 47. as in herbicide production via the chemical synthon processes. Many industrial applications are. 48. currently utilizing valine (Stoner et al., 2000; Eggeling et al., 2001), and thus the global demand. 49. for valine – especially food-grade valine – has grown in the past few years (Cheiljedang, 2013).. 50. Therefore, an intensive study of valine metabolism will be important for gaining a better. 2.

(3) 51. understanding of the metabolic regulation of valine, which may be applicable for both industrial. 52. applications and medical research. Generally, valine is biosynthesized by plants, algae, fungi,. 53. bacteria and archaea, but not in animals. In addition, biotechnology processes have been applied. 54. to industrial valine production in order to meet consumer demand. Microorganisms appear to. 55. be promising hosts for large-scale production, since they can provide high productivity, rapid. 56. growth rates, the ability to utilize several substrates and climate-independent culture processes.. 57. For many years, valine has been commercially produced from bacteria, such as. 58. Corynebacterium glutamicum and Escherichia coli, using several metabolic engineering. 59. approaches to improve the productivity (Park et al., 2007; Park and Lee, 2010).. 60. The yeast Saccharomyces cerevisiae has received an attention as a suitable host for food-. 61. and pharmaceutical-grade products due to its safety and generally recognized as safe (GRAS). 62. status. Yeast cells contain high concentrations of protein, RNA, lipids, amino acids and vitamins. 63. that can be utilized as nutrient-rich dietary supplements for animals and humans. In addition,. 64. yeast extract is a product of the enzymatic digestion of yeast cells, and is currently applied in. 65. several industries – e.g., as a flavoring agent in canned foods and packaged vitamin supplements,. 66. and as a nitrogen source for microorganisms in scientific experiments (York and Ingram, 1996;. 67. Chae et al., 2001). Based on their reliability and safety in food production, the development of. 68. novel yeast strains that overproduce valine may make an important contribution to food-related. 69. industries. Another applicable aspect of valine biosynthesis in S. cerevisiae is isobutanol. 70. production (Hazelwood et. al., 2008). Isobutanol, which is a second-generation biofuel, is. 71. produced from valine degradation in the cytosol, and therefore the enzymes associated with. 72. valine biosynthesis have been engineered to be relocalized from mitochondria to cytosol (Brat. 73. et. al., 2012) (Fig. 1). Based on a screening of butanol-tolerant microorganisms, S. cerevisiae. 74. has been categorized as a better candidate for industrial butanol production than E. coli. 75. (Knoshaug and Zhang, 2009).. 3.

(4) 76. Valine in S. cerevisiae is primarily biosynthesized in the mitochondria from two pyruvate. 77. molecules by mitochondrial enzymes, acetohydroxyacid synthase (AHAS; Ilv2/Ilv6; also. 78. referred to as acetolactate synthase), acetohydroxyacid isomeroreductase (AHAIR; Ilv5),. 79. dihydroxyacid dehydratase (DHAD; Ilv3), to produce the key intermediate, α-ketoisovalerate. 80. (KIV) (Fig. 1). By the assistance of unknown keto-acid transporters, KIV is partially. 81. transported to the cytosol. The final step of valine biosynthesis occurs in both mitochondria and. 82. cytosol via the mitochondrial and cytosolic BCAA aminotranseferase Bat1 and Bat2,. 83. respectively (McCourt and Duggleby, 2005). AHAS is the rate-limiting enzyme for BCAA. 84. biosynthesis and catalyzes the first step of BCAA biosynthesis by converting pyruvate. 85. molecules to 2-acetolactate (Umbarger and Brown, 1958). This enzyme consists of two subunits,. 86. the catalytic subunit (Ilv2) and the regulatory subunit (Ilv6). The expression of the ILV2 and. 87. ILV6 genes is regulated by the general amino acid synthesis activating transcription factor Gcn4. 88. (Magee and Hereford, 1968; Xioa and Rank, 1988; Pang and Duggleby, 1999). In addition, the. 89. enzymatic activity of AHAS is negatively regulated by valine via the process known as. 90. feedback inhibition. A previous study revealed that AHAS activity of the catalytic subunit alone. 91. is unaffected by high concentrations of BCAA, while the enzymatic activity of the reconstituted. 92. catalytic subunit and regulatory subunit were inhibited by valine, indicating a significant role. 93. of the regulatory subunit in feedback regulation (Cullin et al., 1996; Hill et al., 1997; Pang and. 94. Duggleby, 2001). Moreover, other valine derivatives, such as N-acetylvaline, N-methylvaline. 95. and valinamide, had no effect on the AHAS activity, suggesting that valine binds to the. 96. regulatory subunit in a specific manner. Taken together, these results indicate that the AHAS. 97. regulatory subunit Ilv6 is essential for the feedback inhibition by valine and for the full. 98. enzymatic activity (Pang and Duggleby, 2001).. 4.

(5) 99. 100 101. In this study, we focused on the metabolic regulation of valine, particularly the regulatory. 102. subunit of AHAS (Ilv6), in S. cerevisiae. We successfully constructed yeast strains that. 103. significantly increase cellular valine levels by amino acid substitutions in Ilv6 based on a. 104. protein sequence comparison with the AHAS regulatory subunit of E. coli. It was also found. 105. that the Asn86Ala, Gly89Asp and Asn104His variants were less sensitive to feedback inhibition. 106. by valine than the wild-type Ilv6. Finally, we discuss the mechanism of valine overproduction. 107. by the above-described Ilv6 variants. This study reports the removal of feedback inhibition of. 108. AHAS activity, leading to valine overproduction in yeast cells.. 109 110. 2. Materials and methods 5.

(6) 111. 2.1. Strains and culture media. 112. S. cerevisiae BY4741 (derived from S288c) was used in this study. Yeast cells were grown. 113. in a nutrient-rich YPD medium (10 g/L yeast extract, 20 g/L peptone and 20 g/L glucose) for. 114. routine culture. A synthetic dextrose (SD) minimal medium (1.7 g/L yeast nitrogen base without. 115. amino acid and ammonium sulfate, 5 g/L ammonium sulfate and 20 g/L glucose, pH 6.0) was. 116. used for cells harboring pRS416 and pRS415-CgHIS3MET15. The valine toxic analog. 117. norvaline (Chem-Impex International, Inc., Wood Dale, IL, USA) was supplemented at a. 118. concentration of 1 mg/mL or 10 mg/mL for the screening of valine-accumulating cells. An E.. 119. coli DH5α was used for plasmid propagation and the bacterial transformation was carried out. 120. by the method for high efficiency transformation (Inoue et al., 1990). Bacterial cells were. 121. cultured in Luria-Bertani (LB) medium (5 g/L yeast extract, 10 g/L tryptone and 10 g/L NaCl). 122. contained 100 µg/mL of ampicillin. If necessary, 2% agar was added to solidify the medium. In. 123. case of bacterial cells harboring pDONR221, cells were cultured in LB medium containing 50. 124. µg/mL of kanamycin. For protein expression, E. coli Rosetta™ (DE3) pLysS cells were cultured. 125. in LB medium containing 100 µg/mL ampicillin and 30 µg/mL chloramphenicol (Wako Pure. 126. Chemical Industries, Tokyo, Japan).. DL-. 127 128. 2.2. Plasmid construction. 129. Yeast centromere plasmid (YCp) pRS416 was employed for construction of the E. coli-S.. 130. cerevisiae shuttle vectors in this study. The ILV2 and ILV6 genes were amplified from S.. 131. cerevisiae BY4741 genomic DNA by KOD FX DNA polymerase (Toyobo, Osaka, Japan) using. 132. corresponding primers (Supplementary Table S1), and subsequently cut and inserted to pRS416. 133. at the SmaI and NotI recognition sites. Site-direct mutagenesis was used to introduce several. 134. amino. 135. Met276Ala/Met, into the Ilv6 protein. Each mutation was introduced into the ILV6 gene on. acid. substitutions,. Asn86Ala,. Gly89Asp,. 6. Asn104His,. Ile255Ala/Arg. and.

(7) 136. pRS416 using QuikChange II XL Site-Directed Mutagenesis Kit (Agilent, Santa Clara, CA,. 137. USA) and mutagenic primer pairs (Supplementary Table S1). PCR products were then digested. 138. with DpnI before introduction into E. coli DH5α cells as the same manner as described by Inoue. 139. et al. (1990). Plasmid pRS416, which expresses the Ilv6 variants, was further confirmed by. 140. DNA sequencing using BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems,. 141. Waltham, MA, USA). Plasmid pRS415-CgHIS3MET15 (constructed by S. Morigasaki) was co-. 142. transformed with the constructed pRS416-based plasmids to complement the auxotrophic. 143. phenotype of S. cerevisiae BY4741 using high-efficiency transformation method (Burke et al.,. 144. 2000). The plasmids for protein expression were constructed by Gateway® cloning technology. 145. (Invitrogen, Carlsbad, CA, USA), based on the site-specific recombination system. The wild-. 146. type ILV2, ILV6 or mutant ILV6 genes were amplified and all PCR products were tagged with. 147. attB1 and attB2 sequences. PCR products were purified and inserted to the pDONR22 using. 148. BP clonase™ II (Invitrogen, USA). The BP clonase mixture was subsequently introduced to E.. 149. coli DH5α cells and confirmed by DNA sequencing. Desired transformants were subjected to. 150. plasmid DNA extraction and then incubated with protein expression plasmid, pET-53-DEST. 151. using LR clonase™ II (Invitrogen, Carlsbad, CA, USA) and introduced into E. coli DH5α cells.. 152 153. 2.3. Spot test for DL-norvaline resistance. 154. S. cerevisiae cells harboring pRS416-ILV6 series were cultured in 5 mL of SD medium.. 155. After overnight incubation at 30 ºC with rotary shaking, cells corresponding to an OD600 of 10. 156. were collected, washed twice with distilled water and suspended in 1 mL of water. Subsequently,. 157. 10-fold serial dilutions of yeast cells were prepared, 2.5 µL of each dilution was dropped on SD. 158. agar plates containing 1 mg/mL or 10 mg/mL of DL-norvaline and incubated at 30 ºC for 3 days.. 159 160. 2.4. Measurements of intracellular amino acid content 7.

(8) 161. S. cerevisiae cells were pre-cultured in 5 ml of SD medium for 16 h at 30 °C and transferred. 162. to 25 mL of SD medium for cultivation at 30 °C with shaking at 200 rpm. Yeast cells. 163. corresponding to an OD600 of 10 were collected, washed twice with distilled water and. 164. suspended in 500 µL of distilled water. Intracellular amino acids in cell suspension were. 165. extracted by boiling for 15 min at 100 °C. Cell debris was centrifuged at 13,000 × g for 1 min. 166. and subsequently omitted by filtration using 0.2 µm syringe filter (mdi™, Ambala Cantt, India).. 167. Samples were subjected to analysis with an amino acid analyzer (AminoTac™ JL500/V; JEOL,. 168. Tokyo, Japan). Intracellular amino acid concentrations were expressed as a percentage of dry. 169. cell weight.. 170 171. 2.5. Protein expression. 172. E. coli Rosetta™ (DE3) pLysS cells with pET-53-DEST harboring ILV2 and ILV6 series. 173. were pre-cultured at 30 ºC for overnight in LB medium containing 100 µg/mL ampicillin and. 174. 30 µg/mL chloramphenicol. The main culture was performed by inoculating cells into a 300. 175. mL flask at the initial OD600 of 0.008 and incubating at 37 ºC until OD600 was reached to 0.3-. 176. 0.4. Protein expression was induced by adding isopropyl β-D-1-thiogalactopyranoside (IPTG). 177. to a final concentration of 100 µM and cultured at 18 ºC for 18 h with a high aeration rate of. 178. 250 rpm. Cells were harvested by placed on ice for 5 min and centrifuged at 4 ºC for 10 min at. 179. 5,000 × g. Cell pellets were resuspended in ice-cold lysis buffer (500 mM NaCl, 20 mM Tris-. 180. HCl, pH 7.9) containing protease inhibitors.. 181 182. 2.6. Protein purification. 183. All reagents were stored at 4 ºC. Lysozyme (10 mg/g cells) was added to the mixture above.. 184. Cell lysate mixture was subjected to sonication (duty cycle 50%, output = 2, 30 sec/on ice 1. 185. min for 5 cycles) and subsequently centrifuged at 10,000 × g for 20 min at 4 ºC, then filtrated 8.

(9) 186. by 0.45 µM diameter filter. Crude protein was purified by His-accept column (Nacalai Tesque,. 187. Kyoto, Japan) in the same manner that described in a previous study (Pang and Diggleby, 2001).. 188. Purified-proteins were collected in 1.5 mL tubes containing 100 mM dithiothreitol (DTT) and. 189. 0.2 M potassium phosphate (pH 6.0) at the final concentration was immediately added to the. 190. purified Ilv2 to maintain the catalytic activity. According to the previous study, the purified. 191. proteins can be stored as follows. The purified Ilv2 was stored in the elution buffer contains 10. 192. µM FAD and 20% glycerol. For the purified Ilv6, 0.1 M potassium phosphate (pH 7.0) and 20%. 193. glycerol were added to the buffer; small aliquots were kept at -80 ºC. It should be noted that the. 194. catalytic activity was reduced by 80% when the protein was stored overnight at 4 ºC (Pang and. 195. Diggleby, 1999). Proteins were quantified using Bio-Rad Protein Assay (Hercules, CA, USA). 196. and subjected to SDS-polyacrylamide gel electrophoresis.. 197 198. 2.7. Reconstitution of purified AHAS subunits and Assay of AHAS activity. 199. Reconstitution of purified Ilv2 and Ilv6 (wild-type and variants) was carried out as. 200. described before (Pang and Diggleby, 1999; 2001). The reaction was performed at 30 ºC in the. 201. mixture contained 200 mM pyruvate, 1 mM ThDP, 10 mM MgCl2 and 10 µM FAD in 1 M. 202. potassium phosphate buffer (pH 7.0). The enzyme was pre-incubated at 30 ºC for 15 min in 225. 203. µL of the reaction mixture without pyruvate. Then, 25 µL of 2 M pyruvate was added to the. 204. reaction mixture, and further incubated for 20 min at 30 ºC. In case of the experiment for valine. 205. inhibition, various concentrations of L-valine ranging from 0.25 mM to 1 mM were added. 206. before the incubation. Thirty-five µL of 50% (v/v) sulphuric acid was added to stop the reaction,. 207. and further incubated at 60 ºC for 15 min to facilitate the conversion of α-acetolactate into. 208. acetoin. AHAS activity was measured by single-point colorimetry, as described by Singh et al.. 209. (1988). The color was developed by adding 400 µL of 0.5 % (w/v) creatine and 5% (w/v) α-. 210. naphthol (in 4 M NaOH), and then incubated at 60 ºC for 15 min. The color mixture was. 9.

(10) 211. subsequently measured with a spectrophotometer (U-1100 Spectrophotometer, Hitachi, Tokyo,. 212. Japan) at 525 nm, using the data in a reaction without enzyme as a blank. Standard acetoin was. 213. varied from 0.2 µM to 20 µM. One unit of activity was defined as the amount of enzyme. 214. required to produce 1 µmol of α-acetolactate per min under the above conditions. Specific. 215. activity was expressed as enzyme units per mg of catalytic subunit as determined by the. 216. Bradford protein assay (Bio-Rad, Hercules, CA, USA).. 217 218. 3. Results. 219. 3.1. Design of amino acid substitutions of the AHAS regulatory subunit (Ilv6) for valine. 220. overproduction. 221. E. coli and S. cerevisiae have similar mechanisms for regulating valine biosynthesis, in. 222. which AHAS activity is subjected to feedback inhibition by valine. Therefore, the AHAS. 223. regulatory subunit of E. coli (IlvH; AHAS III) was used as a model in this study. A previous. 224. study in E. coli demonstrated that several amino acid substitutions in the IlvH protein confer. 225. resistance to valine feedback inhibition on AHAS activity (Mendel et al., 2001; Kaplun et al.,. 226. 2006). The protein sequence alignment between the E. coli IlvH and the S. cerevisiae Ilv6. 227. revealed a comparable structure at the ACT (aspartate kinase, chorismate mutase and TyrA). 228. domain and the ALS_ss_C (acetohydroxyacid synthase, small subunit and C-terminal) domain. 229. (Fig. 2A). According to this information, we hypothesized that the amino acid residues involved. 230. in feedback inhibition by valine are Asn86, Gly89, Asn104, Ile255 and Met276 of Ilv6 in S.. 231. cerevisiae, which correspond to Asn11, Gly14, Asn29, Leu131 and Val153 of IlvH in E. coli,. 232. respectively (Fig. 2B). It was notable that Asn11, Gly14 and Asn29 are conserved between the. 233. E. coli IlvH and the S. cerevisiae Ilv6. Three mutations, leading to the Asn11Ala, Gly14Asp. 234. and Asn29His substitutions, were isolated from the spontaneous ilvH mutants, while the. 235. Leu131Ala/Arg and Val153Ala/Asp substitutions were identified by PCR-based random. 10.

(11) 236. mutagenesis at the C-terminal region of IlvH (Mendel et al., 2001; Kaplun et al, 2006). Thus,. 237. amino acid substitutions of the S. cerevisiae Ilv6 that correspond to the previously reported. 238. mutations of the E. coli ilvH were introduced as follows: Asn86Ala, Gly89Asp, Asn104His,. 239. Ile255Ala/Arg and Met276Ala/Asp.. 240. 241 242. 3.2. Effects of amino acid substitutions of the AHAS regulatory subunit (Ilv6) on DL-norvaline-. 243. resistance. 244. To identify the mutations in ILV6 that abrogate feedback inhibition by valine, we tested the. 245. growth phenotypes of yeast cells expressing the wild-type and variant Ilv6 on SD medium. 246. containing 1 mg/mL or 10 mg/mL. DL-norvaline,. 11. which is a toxic valine analogue. It was.

(12) 247. expected that valine accumulation would alleviate the cytotoxicity of. 248. might be incorporated into proteins competitively with valine. As shown in Fig. 3A, the. 249. Asn86Ala, Gly89Asp and Asn104His variants of Ilv6 remarkably increased. DL-norvaline. 250. resistance of yeast cells as compared to the wild-type Ilv6, indicating that the. DL-norvaline-. 251. resistant growth phenotype of the variants was dominant. This result was consistent with. 252. previous studies in E. coli (Mendel et al., 2001; Kaplun et al., 2006), suggesting that these. 253. amino acid substitutions in Ilv6 cause valine accumulation in yeast cells. However, the. 254. Ile255Ala/Arg and Met276Ala/Asp variants did not clearly show the DL-norvaline resistance.. 255. The amino acid substitutions in the ALS_ss_C domain of Ilv6 exhibited a different effect from. 256. the corresponding mutations in the E. coli ilvH. Based on these results, it was found that the. 257. Asn86Ala, Gly89Asp and Asn104His substitutions on Ilv6 are responsible for the DL-norvaline-. 258. resistant phenotype in S. cerevisiae.. DL-norvaline,. which. 259. The predicted structure of the yeast Ilv6 (based on PSI-blast and PyMOL) clearly showed. 260. that Asn86, Gly89 and Asn104 are located in the vicinity of the valine-binding site in the ACT. 261. domain (Fig. 4). In addition, we noted that Asn104 binds to valine upon the dimer assembly.. 262. Since the binding of valine to these residues has not been well studied, we further investigated. 263. the effects of different amino acid substitutions at positions 86, 89 and 104 on the DL-norvaline. 264. resistance, based on the side-chain polarity, charge and hydropathy index (Fig. 3B). The results. 265. showed that any substitutions at these positions increased the. 266. same degree. In addition, the combination of Asn86Ala or Gly89Asp with Asn104His did not. 267. show any clear additive effects.. 12. DL-norvaline. resistance to the.

(13) 268. 269. 13.

(14) 270. 271 272. 3.3. Effects of amino acid substitutions of the AHAS regulatory subunit (Ilv6) on intracellular. 273. valine contents. 274. In general, greater resistance to DL-norvaline reflects a higher level of intracellular valine. 275. (Park et al., 2007; 2011). To examine whether the. 276. associated with intracellular valine accumulation, yeast cells expressing the wild-type and. 277. variant Ilv6 were cultivated in SD liquid medium, and the cellular valine levels were examined. 278. (Fig. 5). As we expected, yeast cells expressing the Asn86Ala, Gly89Asp and Asn104His. 279. variants of Ilv6 showed an approximately 4-fold increase in the valine content compared with. 280. that of the wild-type Ilv6. The additive effect of substitution, Asn86Ala/Asn104His and. 281. Gly89Asp/Asn104His, on the valine content was not observed in agreement with the result of. 282. the spot test on. 283. three substitutions, Asn86Ala, Gly89Asp and Asn104His, significantly decreased the. 284. intracellular valine level to that of the wild-type cells via unknown mechanisms. In contrast,. 285. the Ilv6 variants did not have any significant effects on the intracellular level of the other. DL-norvaline-containing. DL-norvaline-resistant. phenotype is. plates (Fig. 3B). Unexpectedly, the combination of. 14.

(15) 286. BCAAs, leucine and isoleucine (Fig. 5). Thus, it was concluded that the amino acid. 287. substitutions at the positions 86, 89 and 104 of Ilv6 specifically elevated the intracellular valine. 288. contents, leading to the DL-norvaline-resistant phenotype of yeast cells.. 289. 290 291. 3.4. Effects of amino acid substitutions of the AHAS regulatory subunit (Ilv6) on enzymatic. 292. properties. 293. Since the Ilv6 variants, Asn86Ala, Gly89Asp and Asn104Ala, exhibited the. 294. norvaline-resistant phenotype and higher intracellular valine contents, we hypothesized that. 295. these amino acid substitutions induce some conformational change of the valine-binding site, 15. DL-.

(16) 296. leading to reduction of the valine-binding affinity. In order to examine this hypothesis, the. 297. AHAS activity was measured using the in vitro reconstituted AHAS, which are comprised of. 298. Ilv2 (the catalytic subunit responsible for catalytic activity) and Ilv6 (the regulatory subunit for. 299. modulating the catalytic subunit) (Pang and Diggleby, 1999). To prepare the recombinant. 300. proteins, Ilv2 and Ilv6 tagged with 6x His at the amino terminus were expressed in the E. coli. 301. Rosetta™ (DE3) pLysS strain. However, due to the insolubilization of the recombinant proteins. 302. in the cell, the putative transit peptides at the amino terminus of Ilv2 (54 amino acids) and Ilv6. 303. (40 amino acids) were removed to increase the protein solubility, since they are not required for. 304. bacterial protein expression (Pang and Duggleby, 1999). As a result, the solubility of Ilv2 and. 305. Ilv6 without transit peptides was remarkably improved for further purification (data not shown).. 306. The AHAS activities of purified-recombinant Ilv2 and Ilv6 were determined based on the. 307. level of acetoin, which is produced from acetolactate, the product from the AHAS reaction. The. 308. AHAS activity of purified Ilv2 was 34.1 U/mg protein, while no AHAS activity was detected. 309. in the reaction containing purified Ilv6 alone (data not shown). In S. cerevisiae, the ILV2 gene. 310. encoding the catalytic subunit Ilv2 was first identified and cloned by complementation of the. 311. ilv2 mutation, which is deficient in the AHAS activity (Poliana, 1984). In addition, the. 312. regulatory subunit of AHAS was later identified based on the similarity of bacterial AHAS. 313. (Oliver et al., 1992). Ilv6 is not involved in the catalytic activity; however, it is responsible for. 314. the feedback inhibition by valine, since the AHAS activity in ∆ilv6 cells was no longer inhibited. 315. by valine (Cullin et al., 1996). With an increase in the protein amount of Ilv6, AHAS activity. 316. was significantly increased until the saturation point, where the concentration of Ilv6 was. 317. around 100-fold that of Ilv2 (Pang and Duggleby, 1999), suggesting that Ilv6 is required to. 318. achieve the full level of AHAS activity. In this study, high AHAS activity was observed in the. 319. reaction containing Ilv2 and Ilv6 in a ratio of 1:100 (125 ± 6 U/mg of protein for the wild-type),. 320. whereas 34 ± 4 U/mg was detected in the reaction containing only Ilv2. Based on these. 16.

(17) 321. preliminary results, it was suggested that purified Ilv2 and Ilv6 were successfully reconstituted. 322. in this experiment.. 323. We next analyzed feedback inhibition of the AHAS activity by valine (Fig. 6). Enzymatic. 324. assays without valine in the reaction mixture revealed that the AHAS activity reconstituted by. 325. the Ilv6 variants (156 ± 11 U/mg for Asn86Ala, 145 ± 8 U/mg for Gly89Asp and 152 ± 6 U/mg. 326. Asn104His) was slightly higher than that from the wild-type Ilv6 (125 ± 6 U/mg). However,. 327. the activity of AHAS reconstituted by the wild-type Ilv6 and Ilv2 was drastically inhibited by. 328. 0.2 mM valine and consequently decreased when the valine concentration was increased,. 329. indicating that the AHAS activity of S. cerevisiae cells is subject to feedback inhibition by. 330. valine; that is, AHAS is the rate-limiting enzyme in the valine biosynthesis of S. cerevisiae cells.. 331. In contrast, the level of AHAS activity from the Ilv6 variants (Asn86Ala, Gly89Asp and. 332. Asn104His) and wild-type Ilv2 was approximately 70-80% even in the presence of 1.0 mM. 333. valine. These results indicate that diminished sensitivity to valine feedback inhibition in AHAS. 334. causes valine accumulation. It was also noted that the activity of AHAS reconstituted by wild-. 335. type Ilv6 and Ilv2 in the presence of 1.0 mM valine (32 ± 5 U/mg) was decreased to a similar. 336. level as observed in single Ilv2 (34 ± 4 U/mg). These results were also consistent with the. 337. previous report that AHAS activity is inhibited by valine via the regulatory subunit Ilv6 (Cullin. 338. et al., 1996).. 17.

(18) 339. 340 341. 4. Discussion. 342. In bacteria, valine is produced from two pyruvate molecules by four enzymes, AHAS,. 343. AHAIR, DHAD and transaminase B (Oldiges et al., 2014). The key enzyme of valine. 344. biosynthesis is AHAS, which is subjected to the feedback inhibition by valine (Pang and. 345. Duggleby, 1999). However, the feedback inhibition of AHAS by valine in C. glutamicum is. 346. weaker than that in E. coli: 50% of the maximum AHAS activity in C. glutamicum remains. 347. even in the presence of 10 mM valine (Elišáková et al., 2005), whereas 80% of the maximum. 348. AHAS activity in E. coli is inhibited by 4.8 µM valine (Mendel et al., 2001). Therefore, the E.. 349. coli AHAS is a suitable model for the study of feedback inhibition by valine. There are three. 350. AHAS isozymes found in E. coli (AHAS I, II and III), whereas only one isozyme exists in S.. 351. cerevisiae. The enzymatic activities of E. coli AHAS I and AHAS III are inhibited in the. 352. presence of valine via the regulatory subunits, IlvN and IlvH, which are responsible for valine 18.

(19) 353. feedback inhibition. Nevertheless, the protein alignment of E. coli IlvN is unique and rather. 354. different from that of yeast Ilv6 unlike that of the E. coli IlvH (PSI-Blast: Altschul et al., 1997).. 355. Accordingly, we employed the regulatory subunit of E. coli AHASIII (IlvH) as a model of the. 356. regulatory mechanism of valine biosynthesis in S. cerevisiae.. 357. Based on their protein sequence alignment, the E. coli IlvH and the S. cerevisiae Ilv6. 358. showed similarly conserved regions at the ACT domain and the ALS_ss_C domain (Fig. 2).. 359. Here, the intracellular valine increased approximately 4-fold when the Ilv6 variants, Asn86Ala,. 360. Gly89Asp and Asn104His, in which amino acid substitutions are located at the ACT domain,. 361. were expressed (Fig. 5). On the other hand, amino acid changes at the ALS_ss_C domain,. 362. Ile255Ala, Ile255Arg, Met276Ala and Met276Asp, did not have any effects on the valine. 363. feedback inhibition. In general, the ACT domain is responsible for the binding to ligands or. 364. small molecules for the regulation of BCAAs biosynthesis, whereas the ALS_ss_C domain was. 365. reported to be involved in binding of other compounds, such as MgATP (Duggleby, 1997; Pang. 366. and Duggleby, 2001). In E. coli, the residues Asn11, Gly14 and Asn29 in IlvH are highly. 367. conserved among several microorganisms, including S. cerevisiae, suggesting that the amino. 368. acid changes at these three positions could lead to the removal of valine feedback inhibition. In. 369. S. cerevisiae, Asn86 and Asn104 are predicted to locate at the putative valine-binding site (Fig.. 370. 4) (Phyre2: Kelley et al., 2015); hence, amino acid substitution at the valine-binding site. 371. probably affects the conformation of this pocket. The valine-inhibitory experiments also. 372. showed that the reconstituted-AHAS with the wild-type Ilv6 was sensitive to feedback. 373. inhibition by 0.2 mM of valine, and the AHAS activity was decreased to the basal level (a level. 374. similar to that by Ilv2 alone) upon an increase in the valine concentration. On the other hand,. 375. the activities of AHAS reconstituted with the Ilv6 variants were slightly decreased even when. 376. the valine concentration was increased up to 1.0 mM. These results supported the hypothesis. 377. that amino acid substitutions within the valine-binding site remove valine feedback inhibition,. 19.

(20) 378. leading to intracellular valine accumulation (Fig. 6). The intracellular valine content (about. 379. 0.4% of dry cell weight) for the engineered strains (Fig. 5) corresponds to ca. 2 g valine per. 380. liter intracellular volume. This concentration is much higher than 1 mM valine tested in the in. 381. vitro assays (Fig. 6). The inhibition of AHAS activity from the wild-type Ilv6 reached the. 382. saturation at 0.4 mM valine (Fig. 6). We have not measured the valine inhibition at the. 383. concentration higher than 1.0 mM, since this range (0.4-1.0 mM) has been indicated to be the. 384. most suitable dose in order to see the difference between the wild-type and variants Ilv6.. 385. It is noteworthy that Asn104 binds to valine upon dimer assembly, whereas Gly89 functions. 386. as a dimer interface between Ilv6 monomers (Fig. 4) (Phyre2: Kelley et al., 2015).. 387. binding site structure is altered, the interaction between the ligand and proteins may be. 388. perturbed. The 3-phosphoglycerate dehydrogenase (3-PGDH) is a good example; this enzyme. 389. also contains the ACT domain that is regulated through feedback inhibition by serine. In the. 390. serine-binding site, His404 acts as the major residue that binds to serine via the side-chains. 391. interaction between Asn406 and Asn424 (Grant et al., 1996; Chipman and Shaanan, 2001). This. 392. binding is thought to be based on polar-polar interaction, since histidine, serine and asparagine. 393. contain a polar side-chain. This assumption is complementary to a previous study in E. coli in. 394. which the amino acid replacement of AHAS IlvH at Gly14 and Asn29 resulted in an inability. 395. to bind with valine, based on the conserved residues of a 3-PGDH model (Grant et Al., 1996;. 396. Mendel et al., 2001). However, the interaction between valine and its binding site in the AHAS. 397. regulatory subunit, Ilv6, is still unclear due to amino acid side-chains of the valine-binding site. 398. associating with polar and non-polar side chains. Meanwhile, the Asn86Ala, Gly89Asp and. 399. Asn104His variants displayed the. 400. which is also located at the ACT domain, was opposed. This variant was isolated as one of the. 401. DL-norvaline-resistant. 402. not only that the genes involved in the BCAA biosynthetic pathway can affect. DL-norvaline-resistant. When the. phenotype; the Val132Ile variant,. mutants from S. cerevisiae S288C (data not shown). This result suggests. 20. DL-norvaline.

(21) 403. resistance by increasing valine productivity, but also that there are some upstream pathways. 404. which play a role in DL-norvaline resistance in a valine-independent manner.. 405. Amino acid substitutions at the carboxyl terminus of Ilv6, Ile255Ala/Arg and. 406. Met276Ala/Asp, in S. cerevisiae did not have any effects on the. 407. intracellular valine, in contrast to the Leu131 and Val153 substitutions in the E. coli IlvH,. 408. respectively. Both Leu131 and Val153 are located in the hydrophobic core of the ALS_ss_C. 409. domain. Leu131 links between the α-4 helix and the β-sheet, and a mutation in this region can. 410. affect the α-4 helix folding. Val153 is located between two monomers and is involved in the. 411. inter-monomer interaction (Fig. 7A) (Kaplun et al., 2006). Homology modeling of the S.. 412. cerevisiae Ilv6 (Fig. 7B), based on the E. coli IlvH, suggests that there are hydrogen bonds. 413. between Ile255, Lys251 and Leu259 that are not directly involved in α-helix packing unlike the. 414. corresponding position on the E. coli IlvH, Leu131, which bound to Thr127, Ser128, Phe134. 415. and Leu153. This evidence supports the hypothesis that amino substitution at position 255 in. 416. Ilv6 is not involved in the valine feedback inhibition in S. cerevisiae. Met275Ala/Asp also. 417. appears not to be involved in valine feedback inhibition unlike Val153 of the E. coli IlvH.. 418. However, there is no clear evidence to support this conjecture, since Met276 in the S. cerevisiae. 419. Ilv6 is also located between two β-sheets and forms two hydrogen bonds between them, similar. 420. to Val153 in E. coli (Fig. 7B). The ALS_ss_C domain of AHAS in S. cerevisiae might be. 421. responsible for another ligand-binding, as described in a previous study, in which this domain. 422. was responsible for the binding of MgATP (Pang and Duggleby, 2001). Taken together, this. 423. evidence is consistent with our findings that the amino acid substitutions at Asn86, Gly89 and. 424. Asn104 removed the valine-feedback inhibition (Fig. 6) and increased the intracellular valine. 425. content by approximately 4-fold (Fig. 5) due to the conformational change in the valine-binding. 426. site.. 21. DL-norvaline. resistance or.

(22) 427. 428 429. The results in Fig. 5 also show that the Ilv6 variants did not confer leucine and isoleucine. 430. accumulation. It is known that leucine and isoleucine biosynthesis originate from the same. 431. pathway as valine biosynthesis, although threonine is the substrate for isoleucine rather than. 432. pyruvate. KIV is the main precursor for both leucine and valine biosynthesis, and thus an. 433. increase in AHAS activity could certainly lead to intracellular accumulation of KIV. However,. 434. in our study, the accumulation of KIV or high AHAS activity did not increase the intracellular. 435. leucine and isoleucine. Leu4, α-isopropylmalate synthase I, is a rate-limiting step of leucine. 436. biosynthesis and subject to feedback inhibition by leucine (Kohlhaw, 2003). In the case of. 437. isoleucine, threonine deaminase (Ilv1) catalyzes the conversion of threonine to α-ketobutyrate,. 438. which is a starting point of isoleucine biosynthesis. It has been reported that this enzyme activity. 439. is inhibited by the presence of isoleucine (Berg et al., 2002). Therefore, neither more KIV 22.

(23) 440. availability nor higher AHAS activity can remove the feedback inhibition caused by leucine. 441. and isoleucine, suggesting that, in S. cerevisiae, leucine and isoleucine feedback inhibitions are. 442. regulated in an AHAS-independent manner. Unfortunately, intracellular leucine and isoleucine. 443. have not been reported in a bacterial valine feedback inhibition-resistant AHAS strain, thus the. 444. relationship between each BCAA feedback inhibition is still unknown. However, the leucine. 445. level in Arabidopsisa thaliana cells with valine feedback inhibition-resistant AHAS increased. 446. by approximately 4-fold, whereas only a 3-fold increase was observed for valine (Chen et al.,. 447. 2010). In comparison, two repeats of the AHAS regulatory domain were observed in A. thaliana. 448. (VAT1), although only a single domain is present in the yeast Ilv6. Moreover, the A. thaliana. 449. AHAS activity is rather inhibited by all BCAAs, especially leucine (Lee and Duggleby, 2001),. 450. suggesting that the regulatory mechanism of BCAA biosynthesis via AHAS is different in each. 451. organism.. 452. Comparing the valine biosynthesis between E. coli and S. cerevisiae, the primary substrate. 453. for AHAS is pyruvate, although the E. coli AHAS III prefers α-ketobutyrate as the first substrate. 454. (Gallop et al., 1990). Hence, in the presence of α-ketobutyrate, AHAS III catalyzes the synthesis. 455. of isoleucine rather than valine (Pátek, 2007). Unlike that in bacteria, the AHAS activity in. 456. yeast is partially inhibited by valine, and after the saturation point, additional increases in valine. 457. concentration do not further inhibit the AHAS activity in yeast (Pang and Duggleby, 1999).. 458. These results suggest that there is a different AHAS regulatory subunit in eukaryote cells.. 459. Nevertheless, the regulation of valine biosynthesis via the AHAS regulatory subunit in S.. 460. cerevisiae has not been fully clarified. The crystal structure of the yeast Ilv6 will be required to. 461. resolve this issue.. 462 463 464. 5. Conclusions An improvement of valine productivity would contribute not only to valine-related. 23.

(24) 465. industries but also to isobutanol production via the Ehrlich pathway (Avalos et al., 2013).. 466. Researchers have attempted to increase intracellular valine levels in yeast by overexpressing. 467. the wild-type genes involved in valine biosynthesis. In this study, we introduced an alternative. 468. strategy by focusing on the feedback inhibition of AHAS activity by valine, which is a rate-. 469. limiting step in the valine biosynthesis of S. cerevisiae. We successfully constructed several. 470. variants of the AHAS regulatory subunit Ilv6 (Asn86Ala, Gly89Asp and Asn104Ala) that were. 471. insensitive to the valine feedback inhibition compared with the wild-type protein by introducing. 472. amino acid substitutions based on previous results in IlvH, a similar protein in E. coli. Moreover,. 473. we found that the intracellular valine content in yeast strains expressing these Ilv6 variants was. 474. approximately 4-fold higher than that of the wild-type strain. These results will contribute to. 475. the development of superior yeast strains for valine and isobutanol overproduction.. 476 477. Conflict of interest. 478. The authors declare there is no conflict of interest.. 479 480. Acknowledgements. 481. We are grateful to Yukiko Sugimoto and Susumu Morigasaki (Nara Institute of Science and. 482. Technology) for their technical assistance. This research was partially supported by Ajinomoto. 483. Co., Inc. to H.T.. 484 485. References. 486. Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., Miller, W. and Lipman, D.J.,. 487. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search. 488. programs. Nucleic Acids Res. 25, 3389-3402.. 489. Avalos, J. L., Fink, G. R., Stephanopoulos, G., 2013. Compartmentalization of metabolic 24.

(25) 490. pathways in yeast mitochondria improves the production of branched-chain alcohols. Nat.. 491. Biotechnol, 31, 335-341. http://dx.doi.org/10.1038/nbt.2509.. 492. Berg, J.M., Tymoczko, J.L., Stryer, L., 2002. Amino acid Biosynthesis Is Regulated by. 493. Feedback Inhibition, in: Biochemistry 5th edition. W.H. Freeman, New York, section 24.3.. 494. Brat, D., Weber, C., Lorenzen, W., Bode, H. B., Boles, E., 2012. Cytosolic re-localization and. 495. optimization of valine synthesis and catabolism enables increased isobutanol production. 496. with. 497. https://doi.org/10.1186/1754-6834-5-65.. the. yeast. Saccharomyces. cerevisiae.. Biotechnol.. Biofuels.. 5,. 65.. 498. Burke, D., Dawson, D., Stearns, T., 2000. Methods in yeast genetics: A Cold Spring Harbor. 499. Laboratory course manual, in: Cold Spring Harbor Laboratory. Plainview, Cold Spring. 500. Harbor Laboratory Press, New York, pp.. 501. Burrage, L. C., Nagamani, S. C. S., Campeau, P. M., Lee, B. H. 2014. Branched-chain amino. 502. acid metabolism: from rare Mendelian diseases to more common disorders. Hum. Mol.. 503. Genet., 23(R1), R1-R8. http://doi.org/10.1093/hmg/ddu123. 504. Chae, H. J., Joo, H., In, M.-J., 2001. Utilization of brewer's yeast cells for the production of. 505. food-grade yeast extract. Part 1: effects of different enzymatic treatments on solid and. 506. protein recovery and flavor characteristics. Bioresour. Technol. 76, 253-258.. 507. http://dx.doi.org/10.1016/S0960-8524(00)00102-4.. 508. CheilJedang,. 2013.. Annual. Report.. 509. http://english.cj.net/company/news/press/press_view.asp?ps_idx=5883andNO=41. 510. (accessed 17.06.04). 511. Chen, H., Saksa, K., Zhao, F., Qiu, J., Xiong, L., 2010. Genetic analysis of pathway regulation. 512. for enhancing branched-chain amino acid biosynthesis in plants. Plant J. 63, 573-583.. 513. http://dx.doi.org/10.1111/j.1365-313X.2010.04261.x.. 514. Chipman, D. M., Shaanan, B., 2001. The ACT domain family. Curr. Opin. Struct. Biol, 11, 694-. 25.

(26) 515. 700. https://doi.org/10.1016/S0959-440X(01)00272-X.. 516. Cullin, C., Baudin-Baillieu, A., Guillemet, E., Ozier-Kalogeropoulos, O., 1996. Functional. 517. analysis of YCL09C: Evidence for a role as the regulatory subunit of acetolactate synthase.. 518. Yeast. 12, 1511-1518.. 519. Duggleby, R. G., 1997. Identification of an acetolactate synthase small subunit gene in two. 520. ukaryotes. Gene. 190, 245-249. http://dx.doi. org/10.1016/S0378-1119(97)00002-4.. 521. Eggeling, L., Pfefferle, W., Sahm, H. 2001. Amino acids. In: Ratledge, C., Bjoern, K. eds. Basic. 522. biotechnology. Cambridge University Press, Cambridge, pp. 281-303.. 523. Elišáková, V., Pátek, M., Holátko, J., Nešvera, J., Leyval, D., Goergen, J.-L., Delaunay, S., 2005.. 524. Feedback-resistant acetohydroxy acid synthase increases valine production in. 525. Corynebacterium. 526. http://dx.doi.org/0.1128/AEM.71.1.207-213.2005.. glutamicum.. Appl.. Environ.. Microbiol.. 71,. 207-213.. 527. Gollop, N., Damri, B., Chipman, D., Barak, Z., 1990. Physiological implications of the substrate. 528. specificities of acetohydroxy acid synthases from varied organisms. J. Bacteriol. 172,. 529. 3444-3449.. 530. Grant, G. A., Schuller, D. J., Banaszak, L. J., 1996. A model for the regulation of D-3-. 531. phosphoglycerate dehydrogenase, a Vmax-type allosteric enzyme. Protein Sci. 5, 34-41.. 532. http://dx.doi.org/10.1002/pro.5560050105.. 533. Hazelwood, L. A., Daran, J.-M., van Maris, A. J. A., Pronk, J. T., Dickinson, J. R. 2008. The. 534. Ehrlich pathway for fusel alcohol production: a century of research on Saccharomyces. 535. cerevisiae Metabolism. Appl. Environ. Microbiol. 74, 2259-2266.. 536. Hill, C. M., Pang, S. S., Duggleby, R. G., 1997. Purification of Escherichia coli. 537. acetohydroxyacid synthase isoenzyme II and reconstitution of active enzyme from its. 538. individual pure subunits. Biochem. J. 327, 891-898.. 539. Hutchison, S. N., Zarghami, N. S., Cusick, P. K., Longenecker, J. B., Haskell, B. E. 1983. The. 26.

(27) 540. effect of valine deficiency on neutral amino acid patterns in plasma and brain of the rat.. 541. J. Nutr., 113, 2164-2170.. 542 543. Inoue, H., Nojima, H., Okayama, H., 1990. High efficiency transformation of Escherichia coli with plasmids. Gene. 96, 23-28.. 544. Kaplun, A., Vyazmensky, M., Zherdev, Y., Belenky, I., Slutzker, A., Mendel, S., Barak, Z.,. 545. Chipman, D. M., Shaanan, B., 2006. Structure of the regulatory subunit of. 546. acetohydroxyacid synthase isozyme III from Escherichia coli. J. Mol. Biol. 357, 951-963.. 547. http://dx.doi.org/10.1016/j.jmb.2005.12.077.. 548. Kelley, L. A., Mezulis, S., Yates, C. M., Wass, M. N., Sternberg, M. J. E., 2015. The Phyre2. 549. web portal for protein modeling, prediction and analysis. Nat. Protocols, 10, 845-858.. 550. http://dx.doi.org/10.1038/nprot.2015.053.. 551 552. Knoshaug, E. P., Zhang, M. 2009. Butanol tolerance in a selection of microorganisms. Appl Biochem. Biotechnol. 153, 13-20.. 553. Kohlhaw, G. B., 2003. Leucine biosynthesis in fungi: entering metabolism through the back. 554. door. Microbiol. Mol. Biol. Rev. 67, 1-15. http://dx.doi.org/10.1128/MMBR.67.1.1-. 555. 15.2003.. 556. Lee, Y. T., Duggleby, R. G., 2001. Identification of the regulatory subunit of Arabidopsis. 557. thaliana acetohydroxyacid synthase and reconstitution with its catalytic subunit.. 558. Biochemistry. 40, 6836-6844. http://dx.doi.org/10.1021/bi002775q.. 559 560 561 562. Magee, P., Robichon‐Szulmajster, H. d., 1968. The Regulation of isoleucine‐valine biosynthesis in Saccharomyces cerevisiae. Eur. J. Biochem. 3, 507-511. McCourt, J. A., Duggleby, R. G., 2005. How an enzyme answers multiple-choice questions. Trends Biochem. Sci. 30, 222-225. http://dx.doi.org/10.1016/j.tibs.2005.03.006.. 563. Mendel, S., Elkayam, T., Sella, C., Vinogradov, V., Vyazmensky, M., Chipman, D. M., Barak,. 564. Z. e., 2001. Acetohydroxyacid synthase: A proposed structure for regulatory subunits. 27.

(28) mutagenesis1.. 565. supported. 566. http://dx.doi.org/10.1006/jmbi.2000.4413.. by. evidence. from. J.. Mol.. Biol.. 307,. 465-477.. 567. Oldiges, M., Eikmanns, B. J., Blombach, B., 2014. Application of metabolic engineering for. 568. the biotechnological production of L-valine. Appl. Microbiol. Biotechnol. 98, 5859-5870.. 569. http://dx.doi.org/10.1007/s00253-014-5782-8.. 570. Oliver, S. G., van der Aart, Q. J., Agostoni-Carbone, M. L., Aigle, M., Alberghina, L.,. 571. Alexandraki, D., et al., 1992. The complete DNA sequence of yeast chromosome III.. 572. Nature, 357, 38-46. http://dx.doi.org/10.1038/357038a0.. 573. Pang, S. S., Duggleby, R. G., 1999. Expression, purification, characterization, and. 574. reconstitution of the large and small subunits of yeast acetohydroxyacid synthase.. 575. Biochemistry. 38, 5222-5231. http://dx.doi.org/10.1021/bi983013m.. 576 577. Pang, S. S., Duggleby, R. G., 2001. Regulation of yeast acetohydroxyacid synthase by valine and ATP. Biochem. J. 357, 749-757. http://dx.doi.org/10.1042/bj3570749.. 578. Park, J. H., Lee, K. H., Kim, T. Y., Lee, S. Y., 2007. Metabolic engineering of Escherichia coli. 579. for the production of L-valine based on transcriptome analysis and in silico gene knockout. 580. simulation.. 581. http://dx.doi.org/10.1073/pnas.0702609104.. Proc.. Natl.. Acad.. Sci.. USA.. 104,. 7797-7802.. 582. Park, J. H., Lee, K. H., Kim, T. Y., Lee, S. Y., 2011. Escherichia coli W as a new platform strain. 583. for the enhanced production of L-valine by systems metabolic engineering. Biotechnol.. 584. Bioeng., 108, 1140-1147. http://dx.doi.org/10.1002/bit.23044.. 585. Park, J. H., Lee, S. Y., 2010. Fermentative production of branched chain amino acids: a focus. 586. on. 587. http://dx.doi.org/10.1007/s00253-009-2307-y.. metabolic. engineering.. Appl.. Microbiol.. Biotechnol.. 85,. 491-506.. 588. Pátek, M., 2007. Branched-chain amino acids, in V. F. Wendisch (Ed.), Amino acid biosynthesis. 589. ~ pathways, regulation and metabolic engineering. Springer Berlin Heidelberg, Berlin,. 28.

(29) 590 591 592 593 594. pp. 129-162. Polaina, J., 1984. Cloning of ILV2, ILV3 and ILV5 genes of Saccharomyces cerevisiae. Carlsberg Res. Commun. 49, 577-584. Singh, B. K., Stidham, M. A., Shaner, D. L., 1988. Assay of acetohydroxyacid synthase. Anal. Biochem. 171, 173-179.. 595. Stoner, E.J., Cooper, A.J., Dickman, D.A., Kolaczkowski, L., Lallaman, J.E., Liu, J.H., Oliver-. 596. Shaffer, P.A., Patel, K.M., Paterson, J.B., Plata, D.J., Riley, D.A., Sham, H.L., Stengel,. 597. P.J., Tien, J.H.J. 2000. Synthesis of HIV protease inhibitor ABT-378 (lopinavir). Org.. 598. Process Res. Dev., 4, 264-269. doi:10.1021/op990202j. 599 600 601 602 603 604. Umbarger, H. E., Brown, B., 1958. Isoleucine and valine metabolism in Escherichia coli VIII. The formation of acetolactate. J. Biol. Chem. 233, 1156-1160. Xiao, W., Rank, G. H., 1988. The yeast ILV2 gene is under general amino acid control. Genome. 30, 984-986. York, S., Ingram, L., 1996. Ethanol production by recombinant Escherichia coli KO11 using crude yeast autolysate as a nutrient supplement. Biotechnol. Lett., 18, 683-688.. 605. 29.

(30) 606. Figure legends. 607 608. Fig. 1. The metabolic pathways of from pyruvate to valine and isobuthanol in S. cerevisiae.. 609. Protein names: Ilv2, the catalytic subunit of acetohydroxyacid synthase (AHAS); Ilv6, the. 610. regulatory subunit. 611. reductoisomerase; (AHAIR) Ilv3, dihydroxyacid dehydratase (DHAD); Bat1, mitochondrial. 612. BCAA aminotransferase (BCAT); Bat2, cytosolic BCAA aminotransferase (BCAT); Aro10,. 613. ketoacid (pyruvate) decarboxylase (KDC); Adh2, alcohol dehydrogenase (ADH).. of acetohydroxyacid synthase (AHAS); Ilv5, acetohydroxyacid. 614 615. Fig. 2. The amino acid sequence alignment of AHAS regulatory subunits, the E. coli IlvH. 616. and the S. cerevisiae Ilv6. (A) The schematic representation of the E. coli IlvH and the S.. 617. cerevisiae Ilv6 amino acid sequences. The ACT domain and the ALS_ss_C domain are orange. 618. and yellow, respectively. (B) The amino acid sequence alignment of the E. coli IlvH and the S.. 619. cerevisiae Ilv6. Residues N11, G14, N29, L131 and V153 in IlvH, where substitutions resulted. 620. in the resistance to feedback inhibition by valine (Kaplun et al., 2006), are indicated by red in. 621. orange and yellow box. The amino acid substitutions in Ilv6 are indicated by arrowhead below. 622. the sequences. The numbers are residue numbers. Identical and similar amino acids in the two. 623. proteins are indicated by the same residue and +, respectively. Dashes indicate the absence of. 624. corresponding amino acid residues at the positions.. 625 626. Fig. 3. Growth phenotypes on. 627. Ilv6. Cell suspensions with 10-fold serial dilutions were dropped (2.5 µL each) on SD agar. 628. medium, which was supplemented with 0.1% allantoin instead of ammonium sulfate as the sole. 629. nitrogen source, in the absence or presence of 10 mg/mL DL-norvaline, then incubated at 30 °C. 630. for 3 days.. DL-norvaline-containing. 30. medium of yeast cells expressing.

(31) 631 632. Fig. 4. Homology modeling of the valine-binding site on the Ilv6 monomer. Based on the. 633. known E. coli IlvH molecular structure, the S. cerevisiae Ilv6 molecular structure was. 634. remodeled using Phyre2 and PyMOL. The positions of amino acid substitutions are indicated.. 635. Red, yellow and green indicate α-helix structure, β-sheet structure and a loop structure,. 636. respectively.. 637 638. Fig. 5. Intracellular branched-chain amino acid contents. Yeast cells were grown in SD. 639. medium (pH 6.0) at 30ºC and collected at log-phase (15 h after inoculation). White, light gray. 640. and dark gray bars represent intracellular valine, leucine and isoleucine contents, respectively.. 641. The values are the means and standard deviations of three independent experiments. Asterisks. 642. indicate statistically significant differences in comparison to the control (wild-type cells. 643. (BY4741) with the empty vector) (Tukey’s test, p < 0.05).. 644 645. Fig. 6. Effects of valine on the AHAS specific activity. In vitro reconstituted complex of wild-. 646. type Ilv2 with wild-type Ilv6 or Asn86Ala, Gly89Asp or Asn104His variants of Ilv6 was used. 647. for the AHAS activity assay. Specific activity was expressed as enzyme units per mg of the. 648. catalytic subunit. The values are the means and standard deviations of three independent. 649. experiments. Asterisks indicate statistically significant differences in comparison to the variants. 650. of Ilv6 (Tukey’s test, p < 0.05).. 651 652. Fig. 7. Amino acid substitution at the ALS_ss_C domain of yeast Ilv6. The ALS_ss_C. 653. domain of the E. coli IlvH (A) and the ALS_ss_C domain the of S. cerevisiae Ilv6 (B).. 654. Homology modeling was illustrated by Phyre2 and Pymol. All colors were given according to. 655. the secondary structure; red, yellow and green indicates a helix structure, a sheet structure and. 31.

(32) 656. a loop structure, respectively. The positions of amino acid substitutions were labeled as the. 657. stick with yellow dashes, which indicate hydrogen bonds.. 32.

(33)